Systems and Methods for Suggesting Headlines

1. A computer implemented method for suggesting headlines, comprising: training a machine learning topic model using search logs and historic click through data to account for the potential of a word in a headline to induce a click; receiving, at a computing device, an input of a headline and an article associated with the headline; determining, by a computing device, a topic associated with the article and headline; determining a plurality of words associated with the topic; inputting each of the plurality of words associated with the topic into the trained machine learning topic model to determine a topical click value for each of the plurality of words; and recommending at least one word of the plurality of words having a higher topical click value compared to other words among the plurality of words.

2. The method of claim 1 , further comprising: inputting at least a portion of the headline and the article into the trained machine learning topic model to determine a word click through rate for at least one word in the portion of the headline and the article.

3. The method of claim 1 , further comprising filtering the search logs and historic click through data to remove common words.

4. The method of claim 1 , wherein the at least one word includes at least one bigram formed of two words.

5. A computer implemented method for evaluating the effectiveness of a headline, comprising: training a machine learning headline click-based topic model using search logs and historic click through data to account for the potential of a word in a headline to induce a click; receiving, at a computing device, a headline and an article associated with the headline; determining, by a computing device, a topic associated with the article and headline; determining a topical click value for at least one word within the headline using the trained machine learning topic model based on trending words within the topic associated with the article and headline; and determining a headline score based on the topical click value of at least one word within the headline, the headline score being indicative of a relative likelihood of the headline being clicked through.

6. The method of claim 5 , further comprising determining at least one recommended word for inclusion in a revised headline and outputting the at least one recommended word, the at least one recommended word resulting in a headline score higher than the determined headline score.

7. The method of claim 5 , wherein determining a topical click value further comprises determining a word click through rate for at least one word in the title.

8. The method of claim 5 wherein the at least one word within the headline comprises at least one bigram.

9. A system for suggesting headline terms, comprising: a processor for executing program code; and non-transitory computer storage memory, storing program code that, when executed on the processor, causes the system to perform a method comprising: implementing a machine learning headline click-based topic model; training the machine learning headline click-based topic model using search logs and historic click through data to account for the potential of a word in a headline to induce a click; receiving an article and a headline associated with the article; determining a topic associated with the article; determining at least one word trending in the topic associated with the article; and determining a word click through rate for the at least one word trending in the topic using the trained machine learning headline click-based topic model.

10. The system of claim 9 , wherein the method further comprises: suggesting at least one word from among the at least one word trending in the topic based on its word click through rate.

11. The system of claim 9 , wherein the at least one word is a word contained in the headline and wherein the method further comprises: calculating a headline click through rate score based on the word click through rate of the at least one word.